Pre-analytical variability characterises effects which are introduced to an analysis by manipulation and storage of a biological sample after taking it ex-vivo, but before actually quantifying the respective analyte. In humans, recent studies demonstrated that pre-analytical factors can be an important confounder for immunoassay measurements of circulating hormones of the GH/IGF-system. In contrast, this topic has rarely been addressed in rodent studies. We therefore now systematically investigated if and how commonly used sample processing and pre-treatment protocols affect measured hormone concentrations of the GH/IGF system in rats. Furthermore, we explored if short term fasting, age and circadian rhythmicity have an impact upon the concentrations of IGF-I, IGFBP-2 and IGFBP-3 in rats. ResultsOn average, concentrations of IGF-I were lower by 9.2% (p<0.01), while those of IGF-II and IGFBP-3 were lower by 24% (p<0.01) in EDTA plasma when compared to plain serum from the same rats. In contrast, concentrations of GH were significantly higher in plain plasma samples compared with serum (+137.8%; p<0.01). Repeated freeze/thaw cycles significantly influenced the measurements of serum IGF-II (+25.9%; p<0.01) and IGFBP-3 (+19.3%; p<0.01) when compared to native serum. Pre-treatment of EDTA plasma with protease inhibitors, or immediate storage of EDTA blood on ice, did not significantly affect the outcome of any measurement. Acidification of plasma samples with HCl led to significantly lower IGF-I in samples (−11.9%, p<0.001) and detection of GH was completely hampered in these samples. With respect to biological variability, age (12-week-old vs. 1-year-old male Wistar rats), but not fasting (up to 18h) or circadian rhythmicity affected circulating concentrations of IGF-I and IGFBP-3. ConclusionPre-analytical variability is a potentially confounding factor which also must be considered in rodent studies when analysing and comparing hormones of the GH/IGF system. If and to what extent a specific pre-analytical procedure affects immunoassay measurements in rodent studies cannot be predicted in advance but rather needs to be tested for each analyte separately.
Read full abstract